Your site is not an island. You need to fit in with the rest of the web. (Keith Instone)

I am a firm believer in standards. My arguments in favor of them are typically two:

1. Somebody else might have thought about this before
2. Users might actually have gotten used to doing things a certain way

As a result, one of the things I love to do on any project is a quick competitive eval. I just go out and see what all our competitors are doing. This gives me a good sense of:

• Whether there are any standards
• How strong they are
• Whether it would make any sense to break them

I’m always amazed, though, at how many people simply can’t wait to break standards. They usually cite innovation, and creative disruption, and whatever buzzword happens to be current.

Actually, I’ve come to the realization, over the years, that this may have more to do with personality than anything. If you’re familiar with Myers-Briggs, I’m a (weak) S. S’s tend to be more practical, down-to-earth, and data-driven. The people who I butt heads with tend to be (strong) N’s. They tend to be more abstract, intuitive, and full of ideas. 

I typically handle my N colleagues by getting them to:

• Focus on higher-level issues, and less on details. For example, radio buttons are a pretty darn good way to tell a user that he needs to make just a single-choice. There’s really no need to reinvent this particular wheel. Developing a wizard to help a user pick the product that’s right for them? Now, there’s something that might actually add some real value.
• Solve real problems rather than simply coming up with random new ideas. A colleague of mine likes to make the distinction between innovation (the former) and mere invention (the latter). 

By the by, before you can solve real problems, you have to identify them. And I’ve always been a big fan of ethnography when it comes to that.

Keith may be most famous for his work in getting going

the IA Summit, UXnet, and World Usability Day

(oh, and those glasses)

One swallow does not a summer make. (Aesop)

This one is actually posted in the observation room of my usability lab. It’s a subtle, tasteful reminder that you might want to consider coming to more than just that one usability test, Mr. or Ms. Team Member. 

Yes, I know you’re busy. Yes, I realize that usability tests are a little bit like baseball games – lots of boredom punctuated by rare, brief flurries of excitement that are easy to miss.

You do realize, though, that that one user you saw may not be totally representative of all 10 I will be bringing in this week, right? In other words, if you happen to be at the very first session, there’s really no need to completely redesign the system (or get all defensive or jump off the roof) before the 10:30.

And when it’s time for the report out, do please be a little circumspect when you’re tempted to talk for 10 minutes about the one user who had that one problem that – hate to break this to you – nobody else actually had. The rest of us did see that person, saw a whole bunch of other people as well, and determined that that original user might just be an outlier.

Yup, that’s right. The rest of the team actually attended most of the tests. And, as a matter of fact, I personally happened to attend all of them. In addition, I was paying attention the whole time as well. Finally, I spent probably twice as much time reviewing my notes, looking at the tapes, trying to figure out what it all meant, and putting it all in a form that you could easily digest.

Now, don’t get me wrong. I’m really happy you could make that particular session. And, no, I am not taking attendance. It’s just that the more users you see, the more you’ll understand and the more refined your subsequent judgments will be. No, you don’t have to attend every darn one. Heck, 3 or 4 might be enough to give you a good idea whether what you’re seeing is representative or not. 

(By the way, I actually have not found this to be a real problem for the actual members of the project team. Interaction designers, information architects, writers and even graphic designers are usually there for the duration. It’s often the managerial or business types who are guilty here. And that’s okay. In general, if these types take my report seriously and act on the findings, it’s not totally essential that they see it with their own eyes.)

And, no, Aesop was not blind. I understand 

the sculptor just had trouble “doing eyes.”

It doesn’t matter how many times I have to click, as long as each click is a mindless, unambiguous choice. (Steve Krug)

If there’s one thing that marketeers understand about the web it’s that the more clicks it takes a user to get to their stuff the worse shape the world is in. To put that as a formula, I guess it would go something like, “clicks = evil.” 

Okay, I’m joking. Seriously, the adage is usually something along the lines of, “If users have to click more than 3 times to get what they’re after, they’ll abandon your site.”

Unfortunately, it’s really not that simple. Analytics show that users click more than 3 times all the time. And surveys and usability tests typically show no decrease in user satisfaction when they do so. In fact, the issue has been tackled by such leading lights as Jakob Nielsen and Jared Spool, neither of whom have found any real correlation between number of clicks and user satisfaction.

What they did find, though, was something a little more interesting. In particular, when the user’s choices are simple and straightforward, users are happy to drill down and click away. It’s when those choices aren’t so straightforward that problems arise.

Here’s how it works ... A straightforward path doesn’t involve much more than clicking. A less straightforward one, though, forces the user to think about each step. Add it all up, and the straightforward but “longer” path may take less time than the “shorter” but more confusing one. And even if it doesn’t, the user’s subjective impression will often make them think the more straightforward path actually took them less time and effort

Jakob Nielsen ties it to foraging theory and something called “information scent.” Like a fox after some rabbits, we users will stick with something as long as we’re pretty sure there’ll be a payoff in the end. If that particular woods or field or website doesn’t seem too promising, though, we’ll likely abandon it for happier hunting grounds. 

Clear labels give us good information scent, encouraging us to keep clicking. Poor labels – even if the “game” happens to be only a click away – unfortunately do not. 

Actually, Steve’s a lot more friendly
than this pic would make him seem

Steve also said:

• "Instructions must die."
• "People often test to decide which color drapes are best, only to learn that they forgot to put windows in the room."
• "Happy talk must die."
• "The point of testing is not to prove or disprove something. It’s to inform your judgement."
• "When fixing problems, always do the least you can."

Help doesn't. (Rolf Molich)

Poor help. It just doesn’t get any respect. 

Whenever I run a test and the user seems ready to give up on a particular task, I always ask what they “would do now.” Most of the time, they say they would call. Some would prefer to chat, something that seems to be a lot more popular these days than in the past. And some, of course, simply want to abandon the task altogether, shut down their device, and make themselves a nice, stiff drink.

What people rarely mention is help. And that’s a shame, because they often have a decent chance of finding an answer there.

But people have also been trained to avoid help like the plague. There was a time – and there are still many systems that follow this model – where help was a completely separate system, identified only by a little help link in the upper right-hand corner.

When users clicked on that link, another window typically covered their screen. From there, they would search or browse for their particular question. So, in other words, total task interruption, major mode switching, and lots of time spent basically starting over from scratch. No wonder people shy away from help.

Given all that, how can we get people to use help again? First, don’t call it help. Second, get it out of the right hand corner of the screen. Interestingly, those two ideas are intimately related. Let me explain …

So, if you take help out of the upper right-hand corner, where are you going to put it? Well, one idea is to make it contextual, to put it on the page itself.  Face it, users typically want to get help on something they’re doing right now. They want to know what to put in that field. They want to know what button to choose. So, why not tell them then and there?

And, if you are doing that, why not go ahead and be more specific about what that help is going to be. Instead of putting “help” next to the field, why not just say, “What can I enter here?” or “Is this secure?” or whatever you think the user’s question might actually be.

A variation on this treatment is FAQs. And these are basically just all the questions the user might have on that particular page, but listed all together. Over the last few years, I’ve found FAQs have tested particularly well.

Context-sensitive and specific is definitely the way to go. Not only may they get the user to actually click on them, but they also might actually help the user as well.

Gag!

Also from Rolf:

• "God is on our side, but we can help him."
• "A successful test is a test that uncovers many problems."
• "Even developers have feelings."

If we build it, they will complain. (John Morris)

Ah yes, transfer users. If you’ve never heard that phrase before, it’s really just a fancy way of identifying users who are moving from one system to another. Upgrading your iPhone? You’re a transfer user. Reacting to that redesign of Facebook? You’re a transfer user. Trying to adjust to that new software program at work that replaced the one you’d been using for 10 years? You’re a transfer user.

And one thing we know about transfer users is that they will squawk. They could be moving from steam-powered mainframes to gestural AI systems that read your brainwaves, and you can be as sure as the dawn that your users will protest, grumble, whine, bleat, carp, cavil, grouse.

Face it, it’s just human nature. People just don’t like change. Now, there are a whole bunch of fancy psychological ideas to support this concept – status quo bias, loss aversion, the endowment effect – but it’s just the way people are.

Think of the all the old idioms that describe just this situation. You can’t teach an old dog new tricks. Old habits die hard. The leopard can’t change his spots. And my particular favorite, Don’t move my cheese! Believe me, I’ve seen a lot of cheese-moving over the years.

How to get around the problem? Jared Spool is a big proponent of incremental change. I can highly recommend his article The Death of the Major Relaunch. 

Having worked on many major corporate rollouts over the years, I can also recommend putting together a serious communication plan. Tell your users what’s happening, when’s it going to happen, and why it’s going to be happen. Tell them these things in several different ways. Tell them these things multiple times.

A final idea is simply to wait. Humans are wonderfully adaptable creatures. That evil abomination that had users storming the corporate gates with torches and pitchforks might well have turned into something that they now can’t live without. Just give it time.

And that’s what I generally tell any team I’m on who is ready to jump off the roof after their relaunch generated less than 100% glowing feedback. Take two weeks or a month or so and look at your feedback again. It’ll probably be just fine. If it’s not, though, that’s when you really need to worry.

Affordances are the baby to skeuomorphism's bathwater. (Dan Wineman)

So, before I do anything else, let me define a couple of terms. First, there’s the ghastly “skeuomorphism.” According to Google, it’s simply an “element of a graphical user interface that mimics a physical object.” They offer examples from note-taking apps, such as “yellow legal pads, squared paper, ring binders, etc." 

Skeuomorphs were at one time very popular parts of UIs. In fact, they were rather central in helping early users move from a world they were familiar with (e.g., the office) to one they were not.

Obviously, they can be overdone. Mimicking rich Corinthian leather for an online address book is a great example of a skeuomorphic design element that really doesn’t add anything, and that probably just gets in the way and looks a little cheesy as well. 

Other examples, however, include rather valuable elements, such as radio buttons, buttons with beveled edges, and tabs. Yes, these things mimic the physical environment. But they also help the user an awful lot. That beveled edge? It says, “Click me.” The radio buttons? They say, “Click one of me.” Tabs? They say, “click me for information on the topic listed.”

In this regard, these elements act, not just as skeuomorphs, but as “affordances” as well. And what, pray tell, are those? According to Don Norman, who first applied the term to software, affordances are “the perceived and actual properties of the thing that determine just how the thing could possibly be used.” Knobs on doors are made for turning. Light switches are made for flicking on or flicking off. Buttons are for pushing. They don’t need instructions. They’re not going to be misused.

So, where’s the problem? You can get rid of the wood grain and still make your buttons look clickable, can’t you? I mean, seriously, can’t you?

Well, according to the proponents of “flat design” (another term), you cannot. They’ve taken the horrors of faux leather and wood grain and mixed them in with the useful things, like buttons that look like buttons, field that look like fields, and tabs that look like tabs. And, unfortunately, those proponents are people with names like Microsoft, and Google, and Apple. Sigh … Hopefully, the pendulum will swing back again one of these days.

So, which light switch would you prefer to use?

The users of your product are experts in what they do and how they do it. (Whitney Quesenbury)

Google isn’t exactly helping me here, but I recall a story about a bunch of philosophers sitting around debating how many teeth hen have. Of course, all of them had great arguments about what that number might be, but none of them actually went outside and checked a chicken.

I’m struck sometimes how much some of my meetings over the years have resembled that conclave of philosophers. In general, the syllogism goes something like this:

1. I’m a user
2. I have teeth
3. All users have teeth

Honestly, sometimes I feel that usability engineers are the only ones who want to go find a chicken and pry open its mouth. Yup, we tend to be the empiricists in this matter.

Now, a lot of this comes from simply doing usability tests. It’s amazing what you can learn about - not only the particular subject matter of the test, but all sorts of things about your system, your users, and people and technology in general. Do that a hundred times a year over so many years and, yup, you do get to know those chickens pretty well.

An even more effective method, though, is through field studies. Now, you may know these as “ethnographic studies” or “contextual inquiry,” but all they really involve is simply watching people doing their thing in their own environment. I’m definitely not a purist when it comes to these things, so I typically have no problem just making them a glorified interview. However you do it, you’re sure to get a lot out of it. And no matter who runs 'em.

Oh, by the way … Hens? They don’t have any teeth. Believe me – I just looked.

(not Whitney, by the way)